yea its so strange because its slowly degraded , even though 4.9 was basic in comparison to 2025 , the performance was a big key to the purchase , im hoping its something easy to fix but i doubt it , however , i do believe alot of recent additions has contributed , so maaybe a cut down version without photogrammetry ,smart hybrid and nodes and see if that helps or makes it stable enough to use at high poly counts 

You can see in the GIF in my first post, that even at 33mill, using a decently large brush radius, that the performance was pretty legit. But, after working on the same model, testing, for more than 10min, performance seemed to degrade nearly to the level seen in your video. 

You need to ask Maxime Clement (Exoside founder) if he has any plan to develop a quad remesher plugin for 3DCoat. https://exoside.com/company/about/

yes, very needed The dev team is working hard to make the full migration to python API, more tutorials will coming after that. About custom GUI widgets... Ctc_nick was doing changes using qt, as you can see on his addon   The floating windows are made with QT. 3DC has native QT, but it is only "wrapped" right now. So the windows do not dock in any panel. 

You can draw strokes to help autoretopo to follow the edgeflow   Hope this help   Try this settings for autoretopo on hard surface models   After autopo, you can mark edges as sharp.   Or when select "Mark Seams", assign Sharp Seams automatically for Unwrap

Thank you for letting me know, so is autotopo intended for organic surfaces and for hard surface manual is the way to go?

It is not possible

Hello, Just wanted to make this request, is it possible to provide more tutorials on the python API? Also to make a sub api that will allow us to make custom gui widgets? Just wanted to put that out there. Thanks --Joshua  

I second this, I would gleefully pay for a 3dcoat version.  And I already have the blender one

are there ways to get auto topo to respect a hard edge in a sculpt? it seems to not want to follow hard edges.   Thanks --Joshua

my tests were using on surface mode , i kept the radius brush the exact same for both and uses the same brushes or like for like , 24 million is quite large but its to show hat 3dcoat 2025 will crash and 4.9 did not and infact kept smooth responsive strokes 

A more extensive test shows some kind of dropoff in performance after sculpting for a while. And you can see examples of using large brush sizes and how it cripples performance. Granted, we really should not be trying to make such edits on a dense mesh like this. It's a bit like trying to drive a nail with a wrecking ball...it's way overkill for such large form changes. That is one reason we have a Multi-Level Resolution featureset. Still, using large brushes shouldn't work that slow on a robust PC. Sphere Performance Test 02.mp4

Right off the bat, I see 2 potential things wrong here. One is that the density of the sphere is rather extreme (for what you are attempting to do), at 24 million, when trying to sculpt large scale edits such as huge extrusions. 24 million is the stage where you are sculpting much smaller details. This sphere should be no more than about 5 mill for such large extrusions. The other problem...at least it's a problem (longstanding issue with both sculpting and Texture Painting) for 3DCoat....is that performance generally sucks when using very large brush radius' like you use in the video. Reduce that radius in half and you will see a big difference. I have asked Andrew many times to please run a diagnostic test on application performance to see what the bottleneck is, when using large brush radius'. It doesn't matter if you have the biggest, fastest computer in the world...3DCoat will continue to struggle mightily with large brush radius'. It's some kind of fundamental, core bottleneck imposed by the application, where it just refuses to apply more pixels in a given radius, no matter how up-to-date your system is. I really wish Andrew and the staff could address this before anything else. Performance is often pretty robust in 3DCoat, but it has a few BIG weaknesses...including scene saving and loading (it was supposed to be faster, recently, but it is still very slow in my experience. I had to wait a few minutes to have this 26mill poly sphere At first, I had no issues, but after more extensive testing, I noticed the performance tended to drop off substantially. I don't know if that is a sign of a Memory leak, or the UNDO buffer getting too large or what...but I recently upgraded my system to an AMD Ryzen 9 9950X (16 core/32 thread) 192GB DDR 5 RAM (running at 5600mhz)...to try and help avoid such performance lags, but again, I find that some performance issues have no bearing on SYSTEM performance. It is something in the code that is holding it back, and has been for over 14yrs.    

1. Surface tools change density until voxels applied. This is sense of surface tools. They are faster, there is no vohelizing. 2. I am on Intel core i9, will seek for rizen... But performance is hard task. Too different brush engines.

The .fbx model is in the exchange folder (after Ptex), it should gave been brought back into Blender as the Per Pixel .fbx file.

nope Applinks is here C:\Users\<username>\Documents\AppLinks and the default 3dcoat docs folder is located on C:\Users\<username>\Documents\3DCoat   check this folders and delete the COAT_USER_PATH variable restart the comp and try again